BIROn - Birkbeck Institutional Research Online

Brewer, Rebecca and Biotti, Federica and Bird, Geoffrey and Cook, Richard(2017) Typical integration of emotion cues from bodies and faces in AutismSpectrum Disorder. Cognition 165 , pp. 82-87. ISSN 0010-0277.

Downloaded from: http://eprints.bbk.ac.uk/21290/

Usage Guidelines:Please refer to usage guidelines at http://eprints.bbk.ac.uk/policies.html or alternativelycontact lib-eprints@bbk.ac.uk.

In press at Cognition

Format: Brief article

Running head: Integration of body and face cues in autism

Typical integration of emotion cues from bodies and faces in

Autism Spectrum Disorder

Rebecca Brewer1, Federica Biotti

2, Geoffrey Bird

3,4,5, & Richard Cook

2

1Department of Psychology; Royal Holloway, University of London

2Department of Psychology; City, University of London

3Institute of Cognitive Neuroscience; University College London

4Experimental Psychology Department; University of Oxford

5MRC Social, Genetic & Developmental Psychiatry Centre; King’s College London

*Corresponding author:

Rebecca.Brewer@rhul.ac.uk

Department of Psychology

Royal Holloway, University of London

Egham Hill, Egham,

Surrey, UK, TW20 0EX

Key words: Autism Spectrum Disorder; Social Perception;

Contextual Modulation; Emotion recognition; Body posture

2

Highlights

Cues to another’s emotion are automatically integrated across the face and body

We tested integration of face and body emotion cues in Autism Spectrum Disorder

Integration of face and body emotion cues in individuals with autism was typical

Reliance on body cues was greatest in those with poor facial emotion classification

3

Abstract

Contextual cues derived from body postures bias how typical observers categorize facial

emotion; the same facial expression may be perceived as anger or disgust when aligned with

angry and disgusted body postures. Individuals with Autism Spectrum Disorder (ASD) are

thought to have difficulties integrating information from disparate visual regions to form

unitary percepts, and may be less susceptible to visual illusions induced by context. The

current study investigated whether individuals with ASD exhibit diminished integration of

emotion cues extracted from faces and bodies. Individuals with and without ASD completed

a binary expression classification task, categorizing facial emotion as ‘Disgust’ or ‘Anger’.

Facial stimuli were drawn from a morph continuum blending facial disgust and anger, and

presented in isolation, or accompanied by an angry or disgusted body posture. Participants

were explicitly instructed to disregard the body context. Contextual modulation was inferred

from a shift in the resulting psychometric functions. Contrary to prediction, observers with

ASD showed typical integration of emotion cues from the face and body. Correlation

analyses suggested a relationship between the ability to categorize emotion from isolated

faces, and susceptibility to contextual influence within the ASD sample; individuals with

imprecise facial emotion classification were influenced more by body posture cues.

4

1. Introduction

The facial expressions of others are a rich source of social information, conveying cues to

affective and mental states. Correct interpretation of facial expressions is therefore important

for fluent social interaction and wider socio-cognitive development (Adolphs, 2002; Frith,

2009). Previous research indicates that facial emotion perception is affected by the context in

which a facial expression is encountered, suggesting that interpretations are informed by our

knowledge and experience (de Gelder et al., 2006; Feldman-Barrett, Mesquita, & Gendron,

2011). Categorization of morphed facial expressions, for example, is biased by the concurrent

presentation of social interactants (Gray, Barber, Murphy, & Cook, 2017) and other non-

interacting faces (Masuda et al., 2008). Perceived facial expression can also be influenced by

affective vocal cues (de Gelder & Vroomen, 2000; Massaro & Egan, 1996), situational stories

(Carroll & Russell, 1996), and visual scenes (Righart & De Gelder, 2008a, 2008b).

A particularly strong form of contextual influence is exerted by body postures (Hassin,

Aviezer, & Bentin, 2013). The same facial expression can vary in appearance when presented

with different bodily expressions; for example, a facial expression may be classified as angry

when presented in the context of a body expressing anger, but disgusted when presented in

the context of a body expressing disgust (Aviezer et al., 2008; Aviezer, Trope, & Todorov,

2012b). These findings imply that the attribution of affective states involves the integration of

emotion cues from across the face and body. The influence of posture contexts is often

automatic; it occurs despite explicit instructions to disregard non-face information (Aviezer,

Bentin, Dudarev, & Hassin, 2011), and modulates early neurophysiological markers of visual

person processing (Meeren, van Heijnsbergen, & de Gelder, 2005).

The present study sought to examine whether observers with Autism Spectrum Disorder

(ASD) integrate emotion cues from body posture contexts when interpreting others’ facial

emotion. ASD is a neurodevelopmental condition associated with social and communication

difficulties, repetitive behaviors and restricted routines (American Psychiatric Association,

2013). There has been considerable interest in the visual perception of individual with ASD

(Dakin & Frith, 2005; Simmons et al., 2009). Observers with ASD may exhibit a local

processing style that hinders their ability to form unified global percepts (Behrmann, Thomas,

& Humphreys, 2006; Happé, 1999; Happé & Frith, 2006). For example, ASD is associated

with good detection of embedded figures, which requires observers to disregard extraneous

5

information present within a complex pattern or scene, to locate a target element (Ropar &

Mitchell, 2001; Shah & Frith, 1983). Those with ASD may also be less susceptible to

context-induced visual illusions than typical individuals (Happé, 1996; Shah, Bird, & Cook,

2016; but see Manning, Morgan, Allen, & Pellicano, 2017) and show reduced global-to-local

interference when responding to (“Navon”) compound letter arrays (Behrmann, Avidan et al.,

2006). Similarly, individuals with ASD may rely less than typical individuals on contextual

cues to distinguish homographs when reading (Frith & Snowling, 1983; López & Leekam,

2003).

Should observers with ASD exhibit a local processing style, their judgements of facial

emotion may be less susceptible to modulation by body posture contexts. Where observed,

aberrant integration of emotion cues from bodies and faces may contribute to difficulties

attributing affective states sometimes seen in this population (Gaigg, 2014). Observers with

ASD and matched control participants were required to classify expressions drawn from a

morph continuum as either ‘Anger’ or ‘Disgust’. Target expressions were either judged in a

no-context baseline condition, in the presence of a task-irrelevant disgusted posture, or a task-

irrelevant angry posture.

2. Method

2.1 Participants

Nineteen individuals with a clinical diagnosis of ASD (three female; Mage = 34.84 years), and

27 individuals with no current or previous clinical diagnosis (eight female; Mage = 33.85

years) took part in the current study. All participants were aged between 18 and 65 years.

Typical participants were recruited from local participant pools populated by university

students and members of the general public. ASD participants were recruited from a database

maintained by the authors. Individuals with ASD were diagnosed by an independent

clinician, and the Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2000) was

used to assess current severity. Autistic traits were also measured in all participants using the

Autism-Spectrum Quotient (AQ; Baron-Cohen, Wheelwright, Skinner, Martin, & Clubley,

2001). Higher AQ scores, indicative of more ASD traits, were seen in the ASD group than in

the typical group [t(44) = 7.28, p < .001]. Alexithymia, a trait associated with difficulties

identifying and describing one’s own emotions (Brewer, Cook, & Bird, 2016a, 2016b),

measured by the Toronto Alexithymia Scale (TAS-20; Bagby, Taylor, & Parker, 1994), was

6

also more severe in the ASD group (M = 59.16, SD = 15.81) than the typical group (M =

46.58, SD = 12.77) [t(43) = 2.95, p = .005]. However, the ASD and typical groups did not

differ significantly in their age [t(44) = .27, p = .787], proportion of female participants [X2 =

.13, p = .248], or IQ [t(43) = 1.032, p = .308]. Detailed diagnostic information is provided in

Table 1.

Table-1

2.2 Stimuli

Facial stimuli (Figure 1a) were static images drawn from a morph continuum created by

blending two images of the same actor expressing disgust and anger (images taken from

Ekman & Friesen, 1975) using Morpheus Photo Morpher Version 3.11 (Morpheus Software,

Inc). The continuum parametrically manipulated the actor’s expression between disgust and

anger in seven equidistant steps of 10%. The body contexts depicted the same actor posing

angry and disgusted postures (Figure 1b). Where these posture contexts have been used

previously (e.g., Aviezer et al., 2008; Aviezer et al., 2012b), the disgusted body posture has

been shown gripping a disgusting object. In the present study, this object was removed to

ensure that perceptual bias, where observed, was attributable to integration of face and body

cues, and not additional semantic information. The morphed facial expressions were

presented within a dark grey oval intended to resemble a hood. The relative location of the

facial target did not vary as a function of expression intensity. In the baseline no-context

condition, observers saw the expressions presented within the oval, but in the absence of a

body posture. We opted for this baseline condition in light of concerns about the value of the

neutral emotion construct (e.g., Lee, Kang, Park, Kim, & An, 2008); for example, a

supposedly “neutral” body posture, where an actor’s arms are clenched by their side, may be

perceived as angry. In all three conditions, facial stimuli subtended 3.5° vertically when

viewed at 57 cm.

Figure-1

2.3 Procedure

Each trial in the experimental procedure began with a central fixation point (1000 ms),

followed by presentation of a facial target drawn from the morph continuum (1200 ms). The

7

facial target was always presented centrally. In the baseline no-context condition, the facial

target was presented in isolation. In the two context conditions, the facial target was

accompanied by an angry or disgusted body posture. Following stimulus offset, participants

were prompted to categorize the facial emotion as either ‘Disgust’ or ‘Anger’ using a key

press. Participants were explicitly instructed to disregard the body context. The procedure

consisted of 420 trials (7 facial stimuli × 20 presentation × 3 context conditions) and was

presented on an LCD display. Stimuli were presented in a randomized order, with the three

context conditions interleaved. The experimental program was written in MATLAB (The

MathWorks, Inc) using the Psychophysics Toolbox (Brainard, 1997; Pelli, 1997).

For each observer, we fitted separate psychometric functions for the three context conditions,

each modelling how the probability of a disgust response varied as a function of the strength

of the disgust signal in the stimulus. Cumulative Gaussian functions were fitted using the

Palamedes toolbox (Prins & Kingdom, 2009). Each function estimated two key parameters:

Decision noise and the point of subjective equality (PSE). Decision noise is a measure of the

precision with which stimuli are categorized, defined as the standard deviation of the

symmetric Gaussian distribution underlying each cumulative Gaussian function. Lower noise

estimates indicate that observers can perceive subtle differences in stimulus strength and vary

their responses accordingly. Greater noise estimates reveal that participants’ responses are

relatively invariant to changes in stimulus strength. Noise estimates are inversely related to

the slope of the psychometric function; steep and shallow slopes are associated with low and

high noise estimates, respectively. The PSE is a measure of bias that represents the

hypothetical emotion intensity equally likely to be judged as ‘Disgust’ and ‘Anger’.

Observers’ susceptibility to the contextual modulation was inferred from the difference

between PSE of the anger function and the PSE of the disgust function (Figure 1c).

3. Results

One individual in the ASD group (participant 19) produced psychometric functions that could

not be modeled so was excluded from all analyses. Goodness-of-fit was evaluated by

calculating deviance scores for each function (McCullagh & Nelder, 1989). In general, fits

were good in both groups; deviance scores (whereby lower deviance indicates better fit) for

those with (M = 6.97, SD = 8.55) and without (M = 8.85, SD = 7.44) ASD did not differ

significantly in the baseline no-context condition [t(43) = .781, p = .439].

8

Figure-2

First, we examined the decision noise of the two groups (Figure 2a). The ASD (M = 15.0%,

SD = 13.9%) and typical (M = 11.7%, SD = 5.94%) groups did not differ significantly with

respect to decision noise in the baseline no-context condition [t(43) = .968, p = .344],

suggestive of broadly comparable emotion recognition. However, the ASD group exhibited

greater variability in decision noise than the typical group (Levene’s F = 9.97, p = .003).

Next, we examined the groups’ susceptibility to the contextual modulation (Figure 2b).

Significant contextual modulation (shifts > 0%) was observed in both the ASD [t(17) = 2.50,

p = .023] and typical [t(26) = 3.43, p = .002] groups. Furthermore, the ASD (M = 7.73%, SD

= 12.9%) and typical (M = 3.47%, SD = 5.26%) groups did not differ in their susceptibility to

the contextual modulation [t(43) = 1.30, p = .207], although modulation was numerically

greater in the ASD group. The ASD and typical groups were also comparable in terms of

their baseline decision noise [t(36) = 1.24, p = .234] and their susceptibility to the contextual

modulation [t(36) = 1.17, p = .261] when participants of 50 years-of-age or over were

excluded (four participants in each group).

Correlation analyses (Figure 2c) conducted on the combined sample of 45 observers revealed

a significant positive association between decision noise in the baseline no-context condition,

and susceptibility to the contextual modulation [r = .63, p < .001]. Participants who exhibited

imprecise classification of isolated facial expressions exhibited greater contextual

modulation. When the ASD and typical groups were separated, this association remained

significant in the ASD group [r = .73, p < .001], even when an outlier with a large context

effect (participant 17 of the ASD sample) was removed [r = .74, p = .001]. This correlation

did not reach significance in the typical group [r = .19, p = .333]. A Fisher r-to-z test

indicated that the strength of the correlation was significantly greater in the ASD than the

typical group [z = 2.22, p = .026].

ADOS score was not significantly correlated with either contextual modulation (r = .03, p =

.901) or decision noise in the baseline no-context condition (r = -.01, p = .983). In the sample

as a whole, AQ scores correlated with context induced PSE shift (r = .32, p = .035), whereby

presence of more ASD traits was associated with greater PSE shifts, but this correlation did

9

not remain significant when participant 17 was removed (r = .18, p = .237). AQ scores were

not associated with decision noise in the baseline no-context condition (r = .15, p = .322).

TAS-20 scores were not significantly correlated with contextual modulation (r = .03, p =

.856) or baseline decision noise (r = -.02, p = .904).

4. Discussion

The current study investigated whether individuals with and without ASD differ in the extent

to which body posture contexts influence interpretation of emotional facial expressions.

Despite explicit instructions to disregard the context, the presence of the body postures biased

observers’ categorization of facial emotion; expressions were more likely to be judged as

angry and disgusted when aligned with angry and disgusted bodies, respectively. Contrary to

prediction, however, participants with and without ASD were influenced to a similar degree

by emotional cues present in the body contexts.

Previous studies indicate that individuals with ASD are less susceptible to contextual effects

in a range of domains (e.g., Behrmann, Avidan et al., 2006; Happé, 1996; López & Leekam,

2003; Ropar & Mitchell, 2001), leading to the view that ASD is associated with difficulties

forming integrated percepts (Behrmann, Thomas et al., 2006; Happé, 1999; Happé & Frith,

2006) and problems using context to inform perceptual inference (Lawson, Rees, & Friston,

2014; Palmer, Lawson, & Hohwy, 2017). Our finding that body contexts bias perception of

facial emotion in observers with ASD is not consistent with this characterisation of ASD. The

reason for this apparent disparity warrants further investigation. In typical observers,

integration of bodily and facial cues likely emerges in response to covariation of facial and

bodily expressions. It is possible that adults with ASD have sufficient visual experience of

these contingencies to develop typical integration mechanisms, despite a local processing

style. Reduced attention to faces (Dawson, Webb, & McPartland, 2005; Klin, Jones, Schultz,

Volkmar, & Cohen, 2002; Nakano et al., 2010) may also promote greater reliance on bodily

cues in this population.

We observed striking variation in expression categorization ability in our ASD sample. While

some members of the ASD group exhibited categorization performance comparable with the

best typical participants, three performed at least two standard deviations below the typical

mean. This adds further weight to the emerging view that deficits of facial expression

10

perception in ASD are weak and inconsistent at the group level (Harms, Martin, & Wallace,

2010; Lozier, Vanmeter, & Marsh, 2014; Uljarevic & Hamilton, 2013). Indeed, one

influential review concluded that “behavioral studies are only slightly more likely to find

facial emotion recognition deficits in autism than not” (Harms et al., 2010, p317). While is it

undeniable that some individuals within the ASD population experience difficulties

recognizing facial emotion, these deficits do not appear to be a universal feature of this

condition, and may instead reflect co-occurring conditions such as alexithymia (Bird & Cook,

2013; Cook, Brewer, Shah, & Bird, 2013).

Interestingly, the susceptibility of those with ASD to contextual modulation was related to

their performance when judging the facial expressions in isolation; observers with imprecise

expression categorization in the no-context condition were influenced more by the body

posture contexts when present. A similar trend was observed in the typical group, however

this correlation did not reach significance, possibly due the limited variability in baseline

decision noise in this sample (e.g., Howitt & Cramer, 2009). While correlations inferred from

small samples must be treated with caution (Schönbrodt & Perugini, 2013), the suggestion

that stimulus ambiguity and contextual influence are closely linked, accords with current

thinking about the role of context in human vision (Bar, 2004; Friston, 2005; Gilbert & Li,

2013; Gregory, 1997). When target stimuli are unmistakable, observers may distinguish

between different perceptual hypotheses without reference to the context. When the physical

attributes of a target stimulus do not clearly distinguish between different perceptual

hypotheses, however, contextual information may be weighted more strongly to resolve the

ambiguity. This view also fits well with the finding that physically similar facial expressions

associated with intense positive and negative emotions (Aviezer, Trope, & Todorov, 2012a),

or created by image morphing (Van den Stock, Righart, & De Gelder, 2007), are particularly

receptive to contextual influence from body postures.

Some readers may be concerned that participants with poor facial emotion recognition chose

to ignore the facial cues entirely, and responded to bodily emotion instead. Crucially, any

attempt to treat the current procedure as a bodily expression recognition task would yield

responses that failed to vary as a function of facial expression intensity. All but one

participant (who was excluded from all analyses), however, produced monotonic

classification functions that could be modelled using a cumulative Gaussian.

11

In conclusion, the current findings suggest that individuals with ASD and typical observers

show broadly comparable integration of emotion cues from faces and bodies. Consistent with

previous studies, individuals with ASD differed widely in their baseline ability to classify

facial expressions. Correlation analyses suggested a relationship between ability to categorise

emotion when faces were presented in isolation, and susceptibility to contextual influence;

those observers with ASD who exhibited imprecise categorization in the baseline condition

were influenced more by body posture contexts. This finding accords with the view that the

visual system weights contextual information strongly when interpreting ambiguous stimuli.

12

References

Adolphs, R. (2002). Neural systems for recognizing emotion. Current Opinion in

Neurobiology, 12, 169-177.

APA. (2013). Diagnostic and Statistical Manual of Mental Disorders (DSM-V). Washington,

DC: American Psychiatric Association.

Aviezer, H., Bentin, S., Dudarev, V., & Hassin, R. R. (2011). The automaticity of emotional

face-context integration. Emotion, 11, 1406-1414.

Aviezer, H., Hassin, R. R., Ryan, J., Grady, C., Susskind, J., Anderson, A., et al. (2008).

Angry, disgusted, or afraid? Studies on the malleability of emotion perception.

Psychological Science, 19, 724-732.

Aviezer, H., Trope, Y., & Todorov, A. (2012a). Body cues, not facial expressions,

discriminate between intense positive and negative emotions. Science, 338, 1225-

1229.

Aviezer, H., Trope, Y., & Todorov, A. (2012b). Holistic person processing: faces with bodies

tell the whole story. Journal of Personality and Social Psychology, 103, 20-37.

Bagby, R. M., Taylor, G. J., & Parker, J. D. A. (1994). The twenty-item Toronto Alexithymia

Scale-II. Convergent, discriminant, and concurrent validity. Journal of Psychosomatic

Research, 38, 33-40.

Bar, M. (2004). Visual objects in context. Nature Reviews Neuroscience, 5, 617-629.

Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001). The autism-

spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning autism,

males and females, scientists and mathematicians. Journal of Autism and

Developmental Disorders, 31, 5-17.

13

Behrmann, M., Avidan, G., Leonard, G. L., Kimchi, R., Luna, B., Humphreys, K., et al.

(2006). Configural processing in autism and its relationship to face processing.

Neuropsychologia, 44, 110-129.

Behrmann, M., Thomas, C., & Humphreys, K. (2006). Seeing it differently: visual processing

in autism. Trends in Cognitive Sciences, 10, 258-264.

Bird, G., & Cook, R. (2013). Mixed emotions: the contribution of alexithymia to the

emotional symptoms of autism. Translational Psychiatry, 3, e285.

Brainard, D. H. (1997). The psychophysics toolbox. Spatial Vision, 10, 433-436.

Brewer, R., Cook, R., & Bird, G. (2016a). Alexithymia: a general deficit of interoception.

Royal Society Open Science, 3, 150664.

Brewer, R., Cook, R., & Bird, G. (2016b). Shared interoceptive representations: The case of

alexithymia. In S. S. Obhi & E. S. Cross (Eds.), Shared Representations:

Sensorimotor Foundations of Social Life (pp. 439-459). Cambridge: Cambridge

University Press.

Carroll, J. M., & Russell, J. A. (1996). Do facial expressions signal specific emotions?

Judging emotion from the face in context. Journal of Personality and Social

Psychology, 70, 205-218.

Cook, R., Brewer, R., Shah, P., & Bird, G. (2013). Alexithymia, not autism, predicts poor

recognition of emotional facial expressions. Psychological Science, 24, 723-732.

Dakin, S., & Frith, U. (2005). Vagaries of visual perception in autism. Neuron, 48, 497-507.

Dawson, G., Webb, S. J., & McPartland, J. (2005). Understanding the nature of face

processing impairment in autism: insights from behavioral and electrophysiological

studies. Developmental Neuropsychology, 27, 403-424.

14

de Gelder, B., Meeren, H. K. M., Righart, R., Van den Stock, J., van de Riet, W. A. C., &

Tamietto, M. (2006). Beyond the face: exploring rapid influences of context on face

processing. Progress in Brain Research, 155, 37-48.

de Gelder, B., & Vroomen, J. (2000). The perception of emotions by ear and by eye.

Cognition & Emotion, 14, 289-311.

Ekman, P., & Friesen, W. V. (1975). Pictures of Facial Affect. Palo Alto, CA: Consulting

Psychologists Press.

Feldman-Barrett, L., Mesquita, B., & Gendron, M. (2011). Context in emotion perception.

Current Directions in Psychological Science, 20, 286-290.

Friston, K. (2005). A theory of cortical responses. Philosophical Transactions of the Royal

Society B: Biological Sciences, 360, 815-836.

Frith, C. (2009). Role of facial expressions in social interactions. Philosophical Transactions

of the Royal Society B: Biological Sciences, 364, 3453-3458.

Frith, U., & Snowling, M. (1983). Reading for meaning and reading for sound in autistic and

dyslexic children. British Journal of Developmental Psychology, 1, 329-342.

Gaigg, S. B. (2014). The interplay between emotion and cognition in autism spectrum

disorder: Implications for developmental theory. Frontiers in Integrative

Neuroscience, 6, 113.

Gilbert, C. D., & Li, W. (2013). Top-down influences on visual processing. Nature Reviews

Neuroscience, 14, 350-363.

Gray, K. L. H., Barber, L., Murphy, J., & Cook, R. (2017). Social interaction contexts bias

the perceived expressions of interactants. Emotion.

Gregory, R. L. (1997). Knowledge in perception and illusion. Philosophical Transactions of

the Royal Society B: Biological Sciences, 352, 1121-1127.

15

Happé, F. (1996). Studying weak central coherence at low levels: children with autism do not

succumb to visual illusions. A research note. Journal of Child Psychology and

Psychiatry, 37, 873-877.

Happé, F. (1999). Autism: cognitive deficit or cognitive style? Trends in Cognitive Sciences,

3, 216-222.

Happé, F., & Frith, U. (2006). The weak coherence account: detail-focused cognitive style in

autism spectrum disorders. Journal of Autism and Developmental Disorders, 36, 5-25.

Harms, M. B., Martin, A., & Wallace, G. L. (2010). Facial emotion recognition in autism

spectrum disorders: a review of behavioral and neuroimaging studies.

Neuropsychological Review, 20, 290–322.

Hassin, R. R., Aviezer, H., & Bentin, S. (2013). Inherently ambiguous: Facial expressions of

emotions, in context. Emotion Review, 5, 60-65.

Howitt, D., & Cramer, D. (2009). Introduction to Research Methods in Psychology. Harlow,

UK: Pearson Prentice Hall.

Klin, A., Jones, W., Schultz, R., Volkmar, F., & Cohen, D. (2002). Visual fixation patterns

during viewing of naturalistic social situations as predictors of social competence in

individuals with autism. Archives of General Psychiatry, 59, 809-816.

Lawson, R. P., Rees, G., & Friston, K. J. (2014). An aberrant precision account of autism.

Frontiers in Human Neuroscience, 8, 302.

Lee, E., Kang, J. I., Park, I. H., Kim, J. J., & An, S. K. (2008). Is a neutral face really

evaluated as being emotionally neutral? Psychiatry Research, 157, 77-85.

López, B., & Leekam, S. R. (2003). Do children with autism fail to process information in

context? Journal of Child Psychology and Psychiatry, 44, 285-300.

16

Lord, C., Risi, S., Lambrecht, L., Cook, E. H., Leventhal, B. L., DiLavore, P. C., et al.

(2000). The autism diagnostic observation schedule-generic: a standard measure of

social and communication deficits associated with the spectrum of autism. Journal of

Autism and Developmental Disorders, 30, 205-223.

Lozier, L. M., Vanmeter, J. W., & Marsh, A. A. (2014). Impairments in facial affect

recognition associated with autism spectrum disorders: a meta-analysis. Development

and Psychopathology, 26, 933-945.

Manning, C., Morgan, M. J., Allen, C. T. W., & Pellicano, E. (2017). Susceptibility to

Ebbinghaus and Muller-Lyer illusions in autistic children: a comparison of three

different methods. Molecular Autism, 8, 16.

Massaro, D. W., & Egan, P. B. (1996). Perceiving affect from the voice and the face.

Psychonomic Bulletin & Review, 3, 215-221.

Masuda, T., Ellsworth, P. C., Mesquita, B., Leu, J., Tanida, S., & Van de Veerdonk, E.

(2008). Placing the face in context: cultural differences in the perception of facial

emotion. Journal of Personality and Social Psychology, 94, 365-381.

McCullagh, P., & Nelder, J. A. (1989). Generalized Linear Models (2nd ed.). New York:

Chapman and Hall/CRC

Meeren, H. K. M., van Heijnsbergen, C., & de Gelder, B. (2005). Rapid perceptual

integration of facial expression and emotional body language. Proceedings of the

National Academy of Sciences of the United States of America, 102, 16518-16523.

Nakano, T., Tanaka, K., Endo, Y., Yamane, Y., Yamamoto, T., Nakano, Y., et al. (2010).

Atypical gaze patterns in children and adults with autism spectrum disorders

dissociated from developmental changes in gaze behaviour. Proceedings of the Royal

Society of London B: Biological Sciences, 277, 2935-2943.

Palmer, C. J., Lawson, R. P., & Hohwy, J. (2017). Bayesian approaches to autism: Towards

volatility, action, and behavior. Psychological Bulletin, 143, 521-542.

17

Pelli, D. G. (1997). The VideoToolbox software for visual psychophysics: transforming

numbers into movies. Spatial Vision, 10, 437-442.

Prins, N., & Kingdom, F. A. A. (2009). Palamedes: Matlab routines for analyzing

psychophysical data. . http://www.palamedestoolbox.org.

Righart, R., & De Gelder, B. (2008a). Rapid influence of emotional scenes on encoding of

facial expressions: an ERP study. Social Cognitive and Affective Neuroscience, 3,

270-278.

Righart, R., & de Gelder, B. (2008b). Recognition of facial expressions is influenced by

emotional scene gist. Cognitive, Affective, & Behavioral Neuroscience, 8, 264-272.

Ropar, D., & Mitchell, P. (2001). Susceptibility to illusions and performance on visuospatial

tasks in individuals with autism. Journal of Child Psychology and Psychiatry, 42,

539-549.

Schönbrodt, F. D., & Perugini, M. (2013). At what sample size do correlations stabilize?

Journal of Research in Personality, 47, 609-612.

Shah, A., & Frith, U. (1983). An islet of ability in autistic children: A research note. Journal

of Child Psychology and Psychiatry, 24, 613-620.

Shah, P., Bird, G., & Cook, R. (2016). Face processing in autism: Reduced integration of

cross-feature dynamics. Cortex, 75, 113-119.

Simmons, D. R., Robertson, A. E., McKay, L. S., Toal, E., McAleer, P., & Pollick, F. E.

(2009). Vision in autism spectrum disorders. Vision Research, 49, 2705-2739.

Uljarevic, M., & Hamilton, A. (2013). Recognition of emotions in autism: a formal meta-

analysis. Journal of Autism and Developmental Disorders, 43, 1517-1526.

Van den Stock, J., Righart, R., & De Gelder, B. (2007). Body expressions influence

recognition of emotions in the face and voice. Emotion, 7, 487-494.

18

Figures

Figure 1

Figure 1. (a) Facial expressions were drawn from a morph continuum that blended anger and disgust emotions

in 10% increments. (b) The facial expressions were either presented in isolation (no context) or aligned with

angry and disgusted body postures. (c) An observer’s susceptibility to the contextual modulation was inferred

from the difference between the PSE of their disgusted function and the PSE of their angry function.

19

Figure 2

Figure 2. (a) Decision noise and (b) PSE estimates for the ASDs and the typical observers in the three

conditions. (c) Scatterplot of the correlation between baseline decision noise and susceptibility to the contextual

modulation.

20

Tables

Table 1. Autism Diagnostic Observational Schedule (ADOS) classification and total score, Autism Quotient

Score, and Full-Scale IQ score for all participants in the ASD group. Mean AQ, IQ, and TAS-20 scores for

typical participants are shown below. Note: one typical individual did not complete the TAS-20, and one did not

complete the IQ assessment.

Participant ADOS Classification ADOS Gender Age AQ Full-Scale IQ TAS-20

1 Autism Spectrum 7 Male 61 45 132 61

2 Autism 10 Male 35 46 112 54

3 Autism 10 Female 22 20 121 44

4 Autism 11 Male 31 37 118 38

5 Autism Spectrum 8 Male 18 21 107 73

6 Autism Spectrum 7 Male 38 40 116 71

7 Autism Spectrum 7 Male 38 17 125 27

8 Autism Spectrum 8 Male 35 36 108 72

9 Autism 14 Male 50 50 121 84

10 Autism Spectrum 9 Male 19 31 92 60

11 Autism 12 Male 31 48 107 64

12 Autism 15 Female 52 45 116 69

13 Autism 15 Male 52 27 118 41

14 Autism 10 Male 21 35 107 72

15 Autism 11 Male 42 33 117 36

16 Autism Spectrum 9 Male 25 39 133 61

17 Autism Spectrum 8 Female 21 46 93 81

18 Autism Spectrum 9 Male 33 35 128 60

19 Autism 14 Male 38 23 78 56

ASD mean

(SD)

34.84

(12.46)

35.47

(10.14)

113.11

(14.03)

59.16

(15.81)

Typical mean

(SD)

33.85

(11.92)

16.93

(5.42)

108.73

(14.06)

46.58

(12.77)